Accelerating the Uptake of 5G for Automotive: Real-World Trials from the TARGET-X Project

As the automotive industry transitions toward high-level autonomy, the demand for connectivity offering deterministic low latency and high reliability becomes paramount. This paper presents the end-to-end design, implementation, and experimental validation of three advanced Vehicle-to-Everything (V2X) use cases within a real-world 5G network environment: Cooperative Perception, Automotive Digital Twin, and Predictive Quality of Service (pQoS) for Tele-operated Driving (ToD). Trials at the 370-hectare IDIADA proving ground in Spain benchmarked 5G and Multi-access Edge Computing (MEC) against 4G and cloud alternatives. Experimental results demonstrate that 5G provides substantial performance gains, achieving average latency reductions up to 90% for V2X messages compared to 4G. The integration of MEC halved the average service latency, consistently maintaining it within the 40–50 ms range required for safety-critical services, whereas cloud-based hosting exhibited uncontrollable fluctuations. In the pQoS for ToD, the implementation of a Network Digital Twin (NDT) and exposure APIs reduced the average video jitter by up to 63%, preventing service collapse in degraded coverage zones. Finally, the automotive Digital Twin (DT) achieved high-fidelity synchronization with temporal deviations consistently below 10%. These findings underscore the necessity of edge-centric architectures and proactive network telemetry for the resilient deployment of future safety-critical mobility services, effectively charting the course for the design of 6G.